Computer Architecture

Lecture 2
Performance
 Defining Performance
• How will you grade one computer better in
performance than the other?
• PCs and Embedded computers
– Response Time
• The time between the start and completion of a task. Includes disk
access, memory access, I/O activities, operating system overhead,
CPU execution time
• Servers
– Throughput( or Bandwidth)
• Number of tasks completed per unit time.
 Check Yourself
• Do the following changes to a computer
system increase throughput, decrease
response time, or both?
– Replacing the processor in a computer with a
faster version.
– Adding additional processors to a system that uses
multiple processors for separate tasks—for
example, searching the web
 Performance (Response Time)
•  
• Performace = 1 / Execution Time
• = =n

• A is ‘n’ times Faster than B

 User CPU Time
• The response time a user experiences also
includes time where OS is sharing CPU for other
programs, and the waiting time for I/O.
• CPU Execution Time is the time spent working
only for the particular task.
• CPU Execution Time
– User CPU Time: Time Spent in the Program
– System CPU Time: Time spent in operating system
working on behalf of program.
System Performance/CPU Performance

• System
  Performance =
• CPU Performance(Our Focus) =
 Example 1
• If computer A runs a program in 10 seconds
and computer B runs the same program in 15
seconds, how much faster is A than B?
 CPU Execution Time and Clock Cycle

•  CPU execution time for a program

= CPU clock cycles * Clock Cycle Time
For a program
• CPU execution time for a program
=
Since clock rate =
 Improving Performance
• Decrease clock cycle time
Or
• Decreased no of clock cycles required by a
program
Decreasing clock cycle time may increase no. of
clock cycles required by a program
 Example 2
• Our favorite program runs in 10 seconds on computer
A, which has a 2 GHz clock. We are trying to help a
computer designer build a computer, B, which will run
this program in 6 seconds. The designer has
determined that a substantial increase in the clock rate
is possible, but this increase will affect the rest of the
CPU design, causing computer B to require 1.2 times
as many clock cycles as computer A for this program.
What clock rate should we tell the designer to target?
 Instruction Performance
• Clock cycle per instruction abbreviated as CPI
is the average no of clock cycles required by
instruction since each instruction may take
different no of clock cycles.
 Example 3
• Suppose we have two implementations of the
same instruction set architecture. Computer A
has a clock cycle time of 250 ps and a CPI of
2.0 for some program, and computer B has a
clock cycle time of 500 ps and a CPI of 1.2 for
the same program. Which computer is faster
for this program and by how much ?
Classic CPU Performance Equation
Example 4
 How to determine the Values
• CPU Execution Time: By running the program
• Clock Cycle Time: Computer Documentation
• Instruction Count: By knowing the ISA. Using
architecture simulator. Does not depend upon
hardware details
• CPI: Depends upon the hardware details
Factors Effecting performance
Check Yourself